用于图像场景分类的空间视觉词袋模型

Bag of Spatial Visual Words Model for Scene Classification

以传统的词袋模型为基础,根据同类场景图像具有空间相似性的特点,提出了一种用于图像场景分类的空间视觉词袋模型。首先将图像进行不同等级的空间划分,针对对应空间子区域进行特征提取和k均值聚类,形成该区域的视觉关键词,进而构建整个训练图像集的空间视觉词典。进行场景识别时,将所有空间子区域的视觉关键词连接成一个全局特征向量进行相似度计算。最终的场景分类结果使用V1滤波器和PACT两种特征在支持向量机LIBSVM上获得。

An approach to recognize scene categories by means of a novel model named bag of spatial visual words was proposed.Images were hierarchically divided into sub regions and the spatial visual vocabulary was constructed by grouping the low-level features collected from every corresponding spatial sub region into a specified number of clusters using k-means algorithm.To recognize the category of a scene,the visual vocabulary distributions of all spatial sub regions were concatenated to form a global feature vector.The classification result was obtained using LIBSVM and two kinds of features were used in the experiments：＂V1-like＂ filters and PACT features.